Page:Illustrations of Indian Botany, Vol. 2.djvu/454

230

ILLUSTRATIONS OF INDIAN BOTANY.

An unlocked for occurrence caused some of the preceding articles to be written more hurriedly than was quite consistent with careful study of the subjects to which they refer. Asclepiadese was one of these. Owing to that circumstance I was at the time of writing prevented satisfying myself of the accuracy of some former observations, regarding which I felt somewhat doubtful. Since the article passed through the press I have been permitted to return to the subject, and now think it advisable to add a brief Postscript appertaining to the tribe Periplocea.

The investigation of the pollen of the members of this tribe is by far the most difficult of the order, but fortunately, an exact knowledge of this part of the organization does not seem so indispensable to the discrimination of its genera, as in the other tribes. In some of the genera I find the pollen loose and granular, even in the anther of the flower-bud, in others it is coherent in the anther but, about the time of expansion of the flower when the anther cells open, it loses its cohesion and becomes attached to the corpuscule in a loose granular form ; while in others, as Hemidesmus and Brachylepis it still presents the form of masses after they have left the anther cell. The circumstance of so many presenting granular pollen after the expansion of the flower, combined with the difficulty of examining it in the early stages of the flower-bud, lead me to the conclusion that an exact knowledge of this part of the organization is less necessary in the limitation of the genera of this tribe than in those of the others. I have not a genuine species of Periploca to examine to ascertain how it is in that genus, but infer, from the circumstance of Decaisne having transferred one of my Streptocaulons to it, that it is the same in both. Finlaysonia I have examined, but scarcely feel satisfied with the result, as regards the pollen, but in all other points it comes so near my Strcptocaulon that I now suspect the latter ought to be reduced and its species referred to the former genus.

As already remarked, Decaisne, following Falconer, however, has transferred one of my species of Slrcptocaulon to Periploca. To this transfer I, with all defference, beg leave to demur, unless they take the rest along with it, and at once reduce my genus. To this I should not object, as I now think the two genera too nearly allied. Strcptocaulon was, in the first instance, constituted to receive a number of Asiatic plants only differino- from Periploca in having glabrous anthers ; those of Periploca being surmounted by a dense tuft of hair. The genus therefore essentially rests on these two circumstances — Asiatic origin and glabrous anthers — Periploca claiming European or African origin, and having hairy anthers.

These characters I do not now think entitled to generic value, and believe that the species of both genera might all be associated under one (with the aid of a section), but so long as Streplocavlon retains a place in the botanical system, I hold that S. calophyllum must remain with it, at all events, cannot be transferred to Periploca, as its admission into that genus breaks down the only characters by which the two are kept distinct, being an Asiatic plant and having glabrous anthers.

I regret not having a specimen of Periploca graca, the type of the genus, with which to illustrate my position by giving a figure of it side by side with Slrept. calophyllum, to show by contrast the difference, but to compensate as far as I can, I give, in an additional plate, generic analyses of that plant and of two species of Streplocaulon, examined by Decaisne and retained by him in that genus, which will show that there is no difference between them and the excluded one ; and in lieu of original dissections, I copy some parts of Lamark's analysis of P. graca.

Having still some room left, I have, in the hope of rendering that supplementary plate more useful, introduced dissections of some other genera of this tribe. These are all more or less perfect, but rather too crowded, being anxious to illustrate as many genera as I could within the small space at my disposal. The subjects introduced into that plate are: three species of Streplocaulon, Periploca grcccu, Finlaysonia, Brachylepis, Decalepis, Cryplolepis, and Cryptostegia. Hemidesmus, Brachylepis, and Decalepis are all nearly allied genera ; Cryptostegia and Cryptolepis are also very nearly allied; while Streptocaulon, Finlaysonia, and Periploca are so much alike that it is now my belief they might be advantageously united.

1. Streptocaulon calophyllum. (R. W.) 6. Decalepis Ilamiltonii. (W. & A.)

2. tomentosum. (R. VV.) 7. Brachylepis nervosa. (W. & A.)

3. extensum. (R. W.) f Cryptolepis Buchanani. (R. B.)

4. Periploca graeca. (Linn.) 9. Cryptostegia grandiflora. (R. B.)

5. Finlaysonia obovata. (Wall.)

PRIM'EJJ BY P. K. HU.M — AMERICAN MISSION PRESS, MADRAS.